This paper is concerned with modelling the stochastic process of income, when its dynamics are given by an aggregate and an individual specific factor. A two-stage methodology is proposed to estimate the parameters of each component. The properties of different estimators are discussed in a rotating panel context, that is, given the number of individuals in each cross-section, the number of periods that an specific individual is observed and the total the number of waves (of the survey). I use this framework to model income dynamics in a Spanish household rotating panel with information on family expenditure. I obtain measures of idiosyncratic and aggregate income risks, which are then used as explanatory variables in a consumption growth equation. This allows me to assess the importance of precautionary saving in response to different sources of income uncertainty.